Somatostatin: A Novel Substrate and a Modulator of Insulin-Degrading Enzyme Activity

Abstract

Insulin-degrading enzyme (IDE) is an interesting pharmacological target for Alzheimer's disease (AD), since it hydrolyzes β-amyloid, producing non-neurotoxic fragments. It has also been shown that the somatostatin level reduction is a pathological feature of AD and that it regulates the neprilysin activity toward β-amyloid.

In this work, we report for the first time that IDE is able to hydrolyze somatostatin [k_cat (s^− 1) = 0.38 (± 0.05); K_m (M) = 7.5 (± 0.9) × 10^− 6] at the Phe6–Phe7 amino acid bond. On the other hand, somatostatin modulates IDE activity, enhancing the enzymatic cleavage of a novel fluorogenic β-amyloid through a decrease of the K_m toward this substrate, which corresponds to the 10–25 amino acid sequence of the Aβ(1–40). Circular dichroism spectroscopy and surface plasmon resonance imaging experiments show that somatostatin binding to IDE brings about a concentration-dependent structural change of the secondary and tertiary structure(s) of the enzyme, revealing two possible binding sites. The higher affinity binding site disappears upon inactivation of IDE by ethylenediaminetetraacetic acid, which chelates the catalytic Zn²⁺ ion. As a whole, these features suggest that the modulatory effect is due to an allosteric mechanism: somatostatin binding to the active site of one IDE subunit (where somatostatin is cleaved) induces an enhancement of IDE proteolytic activity toward fluorogenic β-amyloid by another subunit. Therefore, this investigation on IDE–somatostatin interaction contributes to a more exhaustive knowledge about the functional and structural aspects of IDE and its pathophysiological implications in the amyloid deposition and somatostatin homeostasis in the brain.

Introduction

Insulin-degrading enzyme (IDE, insulysin) is a 110-kDa zinc-metalloprotease, involved in the hydrolysis of short polypeptides that vary significantly in sequence, many of which (such as insulin, β-amyloid, amylin, glucagon, β-endorphin, and atrial natriuretic peptide) show a propensity to form under certain conditions β-sheet-rich amyloid fibrils.¹ IDE expression is ubiquitous in human tissue, being particularly abundant in the brain, liver, and muscles, where it is found primarily in the cytosol, peroxisomes, and endosomes; on the other hand, only a small fraction of the enzyme is located on the plasma membrane and in the mithocondria.2, 3 Genetic studies indicate that IDE region of chromosome 10q is associated to the late-onset Alzheimer's disease (AD) and type II diabetes; furthermore, IDE knocked-out-based work shows glucose intolerance, hyperinsulinemia, and accumulation of β-amyloid in the brain.⁴ These results suggest that IDE may be involved in the pathophysiological pathways common to AD, type II diabetes, and hyperinsulinemia.5, 6, 7

IDE appears to be a multisubunit protein, each subunit being formed by two domains, namely, (i) the N-terminal domain (IDE-N), where the catalytic site is located, and (ii) the C-terminal domain (IDE-C). X-ray crystallography shows that the enzyme looks like a “clam with two valves”, where IDE-N and IDE-C are kept together through a “latch”.⁸ The latch flexibility allows IDE to adopt two different conformations, the “open” and the “closed” state. Only in the open conformation are substrates and reaction products free to go in and out of the active site, favoring the enzymatic activity; on the other hand, in the closed state, the active-site accessibility is severely limited, being characterized by a low enzymatic activity.⁹

Native IDE exists as a mixture of monomer, dimer, and tetramer, which are in equilibrium according to the mass law. The dimeric form has been postulated to be the most active one,¹⁰ although the evidence is quite indirect and not unequivocally proved. This feature might be referable to a greater propensity of the dimeric species to adopt the open conformation. Functional studies on IDE activity modulation by metabolic peptides, such as dynorphins, have suggested an allosteric mechanism, according to which the binding of a peptide to a subunit brings about an alteration of the affinity and enzymatic activity of IDE for substrates interacting with the adjacent one, likely shifting the equilibrium in favor to the open state.¹ However, it must be pointed out that the impossibility to physically separate the oligomeric forms (i.e., the monomer, the dimer, and the tetramer interconnected by the mass law equilibrium) does not allow obtaining distinct functional information on them, vanishing at this stage any attempt to extract meaningful knowledge on the potential cooperative activity of IDE, even though data reported in this article underlie some aspects of the possible intersubunit functional interaction.

IDE substrate specificity has been proposed not to depend on the amino acid sequence (even though it preferentially cleaves basic and hydrophobic amino acids), but mostly on the β-sheet structure recognition. Thus, several peptides cleaved by insulysin adopt the β-sheet conformation when they bind the enzyme, showing a structural arrangement similar to that occurring during self-association of amyloidogenic proteins. Experimental studies confirm that the same residues involved in amyloidogenic protein fibrillation are also responsible for insulysin binding.3, 11

Somatostatin is a cyclic tetradecapeptide first isolated from the hypothalamic tissue as a hormone that inhibits the release of growth hormone.¹² However, it is now considered a multifunctional peptide, located in the central nervous system and in the gastrointestinal system,¹³ where it is involved in the regulation of glucose homeostasis with insulin and glucagon. It has been reported that the somatostatinergic network modulates cognitive and sensory functions in the brain, motor activity, and sleep.¹⁴ In addition, somatostatin level decreases with age,¹⁵ underlying a possible role of somatostatin in the decay of cerebral activities of elder people. A role of somatostatin in the evolution of AD has also been proposed, since the lack of somatostatin in the cortex and hippocampus has been shown to be linked to an impairment of cognitive function and memory.¹⁶ Moreover, a reduction of this neuropeptide has been observed in the cortical and cerebrospinal fluid of AD patients,17, 18, 19, 20 being associated to a selective degeneration of somatostatin-producing neurons21, 22 and an altered expression of all five somatostatin receptors in cortical neurons.¹⁶ Recently, it has been shown that somatostatin regulates β-amyloid metabolism by increasing the enzymatic activity of neprilysin (which is the most important enzyme responsible for the hydrolysis of β-amyloid together with IDE) in primary cortical neurons; a modification of neprilysin localization induced by somatostatin has also been observed.19, 23, 24

Altogether, the available data suggest a possible role of both IDE and somatostatin in the pathogenesis of AD. Therefore, a characterization of somatostatin–IDE interaction and its functional effect on IDE activity should cast some light on the molecular interrelationships at the origin of the pathophysiological events of AD. In the present work, we report for the first time that somatostatin is a substrate of IDE and an allosteric modulator of IDE activity toward a novel fluorogenic β-amyloid (FβA) peptide, establishing the functional basis for a link between IDE activity and somatostatin role in the brain.